sign in sign up
<<
Home , Antifolate, Donald Pinkel, Emil Frei

PERSPECTIVES

but, like surgery, could not eradicate metastatic TIMELINE . Effective treatment for most patients needed to reach every organ in the body. Drugs, biological molecules and immune- and the war on cancer mediated therapies have therefore become the focus for current efforts to cure cancer. From the first experiments with 60 Bruce A. Chabner and Thomas G. Roberts Jr years ago to current attempts to develop drugs for specific cancer-related targets, researchers Abstract | The era of chemotherapy began Of the many challenges of medicine, none has from multiple disciplines have joined together in the 1940s with the first uses of nitrogen had a more controversial beginning and none in the search for more effective cancer drugs. mustards and antifolate drugs. Cancer drug has experienced more hard-fought progress Over time, the development of anticancer development since then has transformed than the treatment and cure of cancer. therapies, based at first on empirical observa- from a low-budget, government-supported Although the neoplastic process has been rec- tions, has become increasingly dependent on research effort to a high-stakes, multi-billion ognized for centuries, little was known about an understanding of human tumour biology. dollar industry. The targeted-therapy the biological mechanisms of transformation revolution has arrived, but the principles and tumour progression until the advent of The first efforts (1940–1950) and limitations of chemotherapy discovered molecular medicine in the latter half of the The beginnings of the modern era of by the early researchers still apply. This twentieth century. Before 1950, therapy chemotherapy can be traced directly to the article chronicles the history of modern remained largely the province of the surgeon. discovery of nitrogen mustard as an effective chemotherapy and identifies remaining Radiation therapy became a valuable tool for treatment for cancer1 (see TIMELINE). In challenges for the next generation of control of local and regional disease after 1960, 1942, Louis Goodman and Alfred Gilman researchers. with the invention of the linear accelerator, and colleagues were recruited by the United

Timeline | The history of chemotherapy

Louis Goodman and Alfred Gilman use The National A combination of Studies by Brian nitrogen mustard to treat Chemotherapy , Druker lead to FDA The FDA approves a patient with non- Program begins at the and The NCI introduces approval of imatinib bevacizumab Hodgkin’s lymphoma and National Cancer The Food and Drug Vincent DeVita (CMF) was ‘disease oriented’ mesylate (Glivec) for (Avastin), the first demonstrate for the first Institute (NCI); a Administration (FDA) and colleagues shown to be effective screening using 60 chronic myelogenous clinically proven anti- time that chemotherapy systematic programme approves the cure lymphomas as adjuvant treatment cell lines derived leukaemia, a new angiogenic agent, can induce tumour for drug screening alkylating agent with combination for node-positive from different types paradigm for targeted for the treatment of regression. commences. cyclophosphamide. chemotherapy breast cancer. of human tumour. therapy in . colon cancer.

1942 1948 1955 1958 1959 1965 1970 1972 1975 1978 1989 1992 2001 2004

Syndey Farber uses Roy Hertz and Min Chiu Li Combination Emil Frei and colleagues The FDA approves The FDA approves Researchers at Harvard University antifolates to demonstrate that chemotherapy demonstrate that for the (Taxol), define mutations in the epidermal successfully induce methotrexate as a single (POMP regimen) is chemotherapy given after treatment of ovarian which becomes the growth factor that confer remissions in agent can cure able to induce long- surgical removal of cancer, a drug that first ‘blockbuster’ selective responsiveness to the children with acute choriocarcinoma, the first term remissions in can would prove to have oncology drug. targeted agent gefitinib, indicating lymphoblastic solid tumour to be cured by children with ALL. improve cure rates activity across a broad that molecular testing might be leukaemia (ALL). chemotherapy. (adjuvant chemotherapy). range of solid tumours. able to prospectively identify subsets of patients that will respond to targeted agents. George Hitchings and Gertude Elion synthesize the 6-.

NATURE REVIEWS | CANCER VOLUME 5 | JANUARY 2005 | 65 PERSPECTIVES

States Department of Defense to examine the Emil Frei and colleagues demonstrated that potential therapeutic value of a series of toxins high doses of methotrexate with leucovorin developed for chemical warfare. In May 1942, prevented recurrence of osteosarcoma follow- Goodman and Gilman, both pharmacologists ing surgical removal of the primary tumour, at the Yale School of Medicine, convinced their establishing the principle of adjuvant collaborator, Gustav Lindskog, a thoracic sur- therapy 8,9.Although this therapy was associ- geon, to treat a patient with non-Hodgkin’s ated with bone-marrow toxicity, the toxic lymphoma with nitrogen mustard2.They effects were reversible, whereas the antitumour proposed that this reagent might destroy a effects cured patients of their cancer. lymphoid tumour, based on autopsy findings The basis for selective effects of these from soldiers dying of exposure to sulphur agents against tumour cells versus normal mustard gas during the First World War. tissue was not apparent from the early labo- These victims had profound lymphoid ratory or clinical experiments. It would take hypoplasia and myelosuppression. Reasoning Figure 1 | Sydney Farber working at his 10 years after the initial studies by Farber that measured doses of a similar agent might microscope. Courtesy of the Dana-Farber and colleagues for Michael Osborn and Cancer Institute, , , USA. cause regression of a lymphatic tumour3, Frank Huennekens to discover, in 1958, that Goodman and Gilman carried out experi- the antifolate drugs specifically inhibi- ments in mice bearing a transplanted lym- ted (DHFR)10,11. phoid tumour. When they observed a marked The antifolates Subsequently, Joseph Bertino (FIG. 2),David level of tumour regression, they convinced A second approach to drug therapy of can- Goldman, Robert Schimke and Bruce Lindskog to inject the closely related com- cer began shortly after the Second World Chabner provided further insight into the pound ‘nitrogen mustard’,a simple but highly War, when Sydney Farber (FIG. 1),a patholo- mechanisms of methotrexate12, leading to reactive molecule, into the bloodstream of a gist at and at the the model for our current understanding of patient with advanced non-Hodgkin’s lym- Children’s Hospital in Boston, investigated the pharmacological principles of cancer phoma and airway obstruction. The medi- the effects of folic acid on patients with chemotherapy. The action of methotrexate astinal and lymphatic masses of the patient leukaemia. This vitamin, which had been depends on active transport into cells regressed. This remission, however, lasted identified by Lucy Wills in 1937 to be the through the reduced- transporter 1 only a few weeks, and disease again pro- factor that was deficient in patients with (RFT-1), its conversion to a long-lived intra- gressed, but the principle was established that megaloblastic anaemia4,seemed to stimu- cellular polyglutamate, and its binding to drugs could be administered systemically to late proliferation of acute lymphoblastic DHFR, which leads to inhibition of the syn- induce tumour regression. leukaemia (ALL) cells when administered thesis of thymidylate and purines and the The same scientists next pursued studies to children with this cancer. Farber’s collab- induction of apoptosis (FIG. 3). to define the molecular action of the mus- oration with Harriett Kilte and the medici- Cellular defects in any of these steps can tard compound, demonstrating its forma- nal chemists at Lederle Laboratories led lead to . Mutations in RFT-1, tion of an alkylating intermediate, the to the synthesis of folate analogues — amplification or mutation of DHFR, loss of ethyleneimmonium ring, which reacted first and then amethopterin polyglutamation, and defects in the apop- with electron-donating sites on proteins (methotrexate) — which Farber adminis- totic pathway have all been shown to lead to and nucleic acids. The principle was estab- tered to children with ALL in the late 1940s loss of efficacy13,14.Methotrexate was also lished that tumours might be more suscep- (REF. 5).By blocking the function of folate- the first drug for which pharmacokinetic tible to toxins than normal tissues, requiring , these agents became the analysis was routinely used to monitor drug although the reasons for this were not first drugs to successfully induce remission clearance and identify patients at risk of understood. The discovery that the reagent in children with ALL. Remissions were severe toxicity15.Methotrexate is still pri- formed a covalent bond with DNA was brief, but the principle was clear — antifo- marily used to treat patients with ALL, as made through later studies that demon- lates could suppress proliferation of malig- well as those with certain lymphomas, strated specific sites of alkylation on purine nant cells, and could thereby re-establish osteosarcoma and choriocarcinoma. The bases, leading to crosslinking of strands and normal bone-marrow function. induction of apoptosis. Other improved As a single agent, methotrexate proved to alkylating agents, developed in the follow- have antitumour activity in a range of epithe- ing 20 years, were chemically stabilized lial malignancies, including breast, ovarian, through electron-rich substitutions, and bladder, and head and neck . However, could be administered orally. Cyclophos- its most remarkable effects were recognized in phamide, and others became two uncommon tumours. In 1958, 8 years after standard components of regimens used to Farber’s discovery of antifolates, Roy Hertz and treat patients with lymphomas, leukaemias Min Chiu Li at the National Cancer Institute and, to a limited extent, solid tumours. (NCI)6 found that methotrexate treatment Unfortunately, Goodman and his collabora- alone could cure choriocarcinoma7,a germ-cell tors noted in their earliest experiments that malignancy that originates in trophoblastic tumours quickly became resistant to these cells of the placenta. This was the first solid Figure 2 | Mentor, Joe Bertino, and student, drugs — an observation that predicted tumour to be cured by drug therapy in Bruce Chabner, at in 1970 with the clinical experience with single-agent humans. Further usefulness of methotrexate Barbara Morrison. Reproduced with permission nitrogen mustards. was demonstrated 16 years later, in 1974, when from REF.90 © AlphaMed Press (2001).

66 | JANUARY 2005 | VOLUME 5 www.nature.com/reviews/cancer PERSPECTIVES

MTX In the 1960s, seminal experiments of Frank Schabel and Howard Skipper at the b MTX Southern Research Institute complemented the NCI’s work, by forming an intellectual framework for analysing the kinetics of a tumour growth, as well as the creation of MTX in vivo assays for quantifying cytotoxicity24,25. c They showed that each dose of anticancer MTX(Glu)n drug therapy killed a fraction of tumour cells. d Depending on the drug, cell killing could FH4 f FH2 CH2FH4 Purine biosynthesis require exposure of cells during a particular stage in the . Inhibitors of DNA syn- dUMP RNA e thesis, such as arabinoside and TTP DNA methotrexate, were most effective against TMP f rapidly dividing cells, whereas drugs that physically damaged DNA, such as alkylating Figure 3 | Mechanism of action of methotrexate. Methotrexate (MTX) enters the cell through the agents, killed cells in all phases of the cell reduced folate carrier (a) using an endocytic pathway activated by a folate receptor (b). After entering the cell, methotrexate is polyglutamated (Glu) by the the folylpolyglutamate synthase (c). Methotrexate cycle. These researchers also showed that and its polyglutamates inhibit the enzyme dihydrofolate reductase (d), thereby blocking the conversion of cytotoxicity was a direct function of dose, dihydrofolate (FH2) to tetrahydrofolate (FH4). As tetrahydrofolate stores are depleted, thymidylate (TMP) and demonstrated the effectiveness of combi- synthesis (e) is reduced,which ultimately inhibits DNA synthesis (f). Long-chain polyglutamates of MTX nation therapies in preventing drug resis- have the same affinity as MTX for the target enzyme dihydrofolate reductase, but have markedly increased tance. Finally, Schabel and Skipper were inhibitory effects on both thymidylate synthesis (e) and purine biosynthesis (f), which is required for RNA the first to suggest that high-dose chemother- production. Adapted from figure 1 of REF. 91. apy might be used to cure patients with oth- erwise refractory tumours. Their work led directly to the current practice of using high- lessons learned from studies of this drug way to prevent drug-resistant tumour cells. dose chemotherapy, along with bone- have provided a valuable model for under- From the 1970s to the present, researchers at marrow transplantation, to treat patients standing mechanisms of resistance for The St. Jude’s Children’s Research Hospital in with lymphoma and leukaemia. other agents. Memphis, Tennessee, under the leadership of Donald Pinkel, Joseph Simone and their Combination chemotherapy Beginnings of modern chemotherapy successors, developed and refined curative Clinically, the move to combination Other antileukaemic drugs came to clinical therapy for ALL. chemotherapy began with the development of trials in the early 1950s through the work of curative therapy for childhood ALL by George Hitchings and Gertrude Elion, who A national treatment research effort Holland, Frei and Freireich. This approach was studied purine analogues such as 6-mercap- This success in treating ALL led to legislation extended to the lymphomas in 1963 by topurine (6-MP)16,17.In synthesizing 6-MP, by the Congress to create a Vincent DeVita, George Canellos and col- Elion and colleagues demonstrated that small National Cancer Chemotherapy Service leagues at the NCI (FIG. 4),who ultimately changes in a compound needed by cells Center (NCCSC) at the NCI in 1955, the proved in the late 1960s that nitrogen could inhibit the growth of tumour cells in first federal programme to promote drug mustard, , and pred- part through the de novo inhibition of early discovery for cancer. As most pharmaceuti- nisone — known as the MOPP regimen — steps preceding RNA and DNA synthesis. cal companies were not yet interested in could cure patients with Hodgkin’s lymphoma Furthermore, the Eli Lilly natural products developing anticancer drugs, the NCCSC and non-Hodgkin’s lymphoma26,27.Other group found that Vinca alkaloids, originally established all the necessary components for promising reports of the ability of combina- discovered in a screen for antidiabetic agents, the discovery, development, toxicological tion chemotherapy to cure diffuse large-cell blocked proliferation of tumour cells18.The testing and clinical evaluation of candidate lymphomas were offered by Joseph Bertino antitumour effect of the Vinca alkaloids was agents. Novel components of the NCCSC and colleagues at Yale University28,29.As pre- later shown to be due to their ability to included the development of animal models dicted by studies in animal models, drugs were inhibit polymerization, and of cancer, the P388 and L1210 lymphoid most effective when used in patients with therefore cell division19.Finally, James leukaemia cell lines, and a range of trans- tumours of smaller volume, and as combina- Holland, Emil Freireich, and Emil Frei plantable solid tumours21–23.Subsequently, tion therapies. Even modestly effective drugs showed in 1965 that a combination of the clinical programmes of the NCCSC such as 5-fluorouracil, an inhibitor of DNA methotrexate (an antifolate), vincristine (a were re-shaped into the Cancer Therapy synthesis, could improve survival when used Vinca alkaloid), 6-MP and prednisone — Evaluation Program (see online links box), as an adjuvant in treating patients with colon which together was referred to as the POMP which continues to have a crucial role in the cancer30.Similarly, the landmark trials of regimen — could induce long-term remis- development of cancer drugs; and the pre- Bernard Fisher, who chaired the National sions in children with ALL20.In a manner clinical efforts fell under the Developmental Surgical Adjuvant Breast and Bowel Project similar to that of therapy for tuber- Therapeutics Program. Paul Carbone and from 1967 to 1994, and of Gianni Bonadonna culosis and subacute bacterial endocarditis, Marvin Zelen of the NCI led the efforts to proved that adjuvant chemotherapy after combinations of drugs, each with a different establish broad-based cooperative group complete surgical resection of breast tumours site of action, proved to be the most effective trials for the common solid tumours. significantly extended survival — particularly

NATURE REVIEWS | CANCER VOLUME 5 | JANUARY 2005 | 67 PERSPECTIVES

be consistently effective in the treatment of reaction at the O6 position of . This patients with ovarian cancer31.Although no drug proved modestly effective in treat- patent position had been established, the ing patients with malignant gliomas42,43. NCI signed a collaborative research and Montgomery’s group also developed flu- development agreement under which it darabine phosphate, a purine analogue, agreed to share clinical and manufacturing which has become a mainstay in treatment data exclusively with Bristol Myers Squibb of patients with chronic lymphocytic in 1991. Taxol subsequently became BMS’s leukaemia44.Other effective molecules came first billion dollar per year drug, leading to from industry during the period of 1970 acrimonious debate in Congress about to 1990, including and industrial exploitation of a government epipodophyllotoxins — both of which Figure 4 | The ‘gang of five’. Left to right: George discovery32 and to calls for price controls. inhibited the action of topoisomerase II45,an Canellos, Bruce Chabner, Phillip Schein, Vincent Another drug class that had a difficult enzyme crucial for DNA replication, tran- DeVita and Robert Young (1970). Photograph start was the . , scription and repair. Additionally, several from the author’s collection. derived from a Chinese ornamental tree, agents originally developed as anticancer inhibits topoisomerase I, an enzyme that drugs proved crucial in the treatment of allows DNA unwinding and strand passage. non-neoplastic diseases, such as methotrex- in those women whose cancer had spread to Despite showing promise in preclinical ate in , cyclophos- the axillary lymph nodes6.In general, combi- studies, the agent had little antitumour phamide in Wegener’s granulomatosis, and nations of drugs proved to be more effective activity in early clinical trials and was toxic nucleoside inhibitors in HIV/AIDS. than single agents against both metastatic can- to kidneys. Its ineffectiveness in vivo was Throughout the clinical development of cer and in patients at high risk of relapse after determined to be due to instability of its lac- anticancer drugs, researchers repeatedly primary surgical treatment. tone ring at neutral pH. When it entered the encountered significant problems because urine, which has an acidic pH, the active of the acute and long-term toxicities of Natural products: trials and tribulations molecule reformed, causing renal tubular , which affected virtually In 1956, C. Gordon Zubrod, who had for- damage. Not until 1996 would a stable every organ of the body. Oncologists merly led the development of antimalarial camptothecin analogue, , finally accepted these as the price for controlling a agents for the United States Army in win Food and Drug Administration (FDA) fatal disease. The lethality of bone-marrow Aberdeen, Maryland, assumed leadership of approval for the treatment of colon cancer33. suppression was significantly ameliorated by the Division of Cancer Treatment of the NCI, Later, this agent would also be used to treat the development of platelet transfusion by and guided development of new drugs at lung and ovarian cancers 34,35. Freireich and colleagues at the NCI46,47,by both the experimental and, later, clinical level. aggressive use of to prevent In the two decades that followed, the estab- NCI’s successes in drug discovery infections in neutropaenic patients48, and by lishment of the NCCSC, a large network of During the 1970s and 1980s, the NCI and the later discovery of growth factors such as cooperative clinical trial groups evolved others in academia, both in the United granulocyte colony-stimulating factor and under the auspices of the NCI to test anti- States and abroad, continued to dominate granulocyte–monocyte colony-stimulating cancer agents, first in children with ALL, and cancer drug development, primarily because factor, which allowed rapid restoration of later in adults with solid tumours. Zubrod there were few participants from industry. neutrophils49,50.Despite the advent of these had a particular interest in natural products, Cancer drug discovery had gained a reputa- supportive measures, the potential of some and established a broad programme for the tion for having high risk and little chance of cytotoxic drugs to cause leukaemia51, as well collecting and testing of plant and marine efficacy. Fewer than 10% of new drugs as their long-term effects on the lungs, heart sources, a controversial programme that led entering clinical trials in the period from and reproductive organs, remain formidable to the discovery of (in 1964) and 1970 to 1990 achieved FDA approval for barriers, and have become increasingly camptothecins (in 1966). marketing36, and animal models seemed important as patients are cured of their Both classes of drug, which were isolated unreliable in predicting clinical success37. primary tumours52. and characterized by the laboratory of However, to be fair, there were successes. Monroe Wall at the Research Triangle Cisplatin, discovered by a Michigan State The need to change strategies Institute, encountered significant problems University researcher, Barnett Rosenberg38, In the early 1980s, progress in chemother- in development. Paclitaxel (Taxol), a novel who was working on an NCI contract, was apy of cancer seemed slow, and each small antimitotic that promoted microtubule instrumental in the cure of testicular cancer 39. success required large, long-term trials that assembly, proved difficult to synthesize and Subsequently, Eve Wiltshaw, Hillary Calvert only led to marginal gains against solid could only be obtained from the bark of the and others at the Institute of Cancer Research tumours. Furthermore, mouse models of Pacific Yew tree, which forced the NCI into in the United Kingdom extended the clinical leukaemia and solid tumours, which were the costly business of harvesting substantial usefulness of the platinum compounds with the mainstays for drug screening at the quantities of yew trees from public lands. their development of carboplatin40,a cisplatin time, were poor predictors of clinical out- Furthermore, this drug was virtually insolu- derivative with broad antitumour activity and come. By 1985, the NCI drug development ble and had to be formulated in a lipid comparatively less nephrotoxicity. A second effort had begun to produce a monotonous emulsion that causes hypersensitivity reac- group with an NCI contract, led by John group of , alkylators, antim- tions in some patients. After 4 years of clini- Montgomery, at the Southern Research itotics and topoisomerase inhibitors. cal testing in solid tumours, it was found in Institute, synthesized nitrosoureas41,which Analogues of these drugs, which provided 1987 (23 years after its initial discovery) to alkylated and crosslinked DNA in a novel marginal increases in efficacy, evoked little

68 | JANUARY 2005 | VOLUME 5 www.nature.com/reviews/cancer PERSPECTIVES

Box 1 | Screening strategies such as Stanley Cohen and Rita Levi- Montalcini identified specific growth factors. Systematic drug screening began in 1955 at the National Cancer Institute (NCI) with the Others deciphered the network of signalling 81 establishment of the Cancer Chemotherapy National Service Center screening programme . molecules that connected these receptors to Throughout the 1960s and 1970s, most screening was performed in vivo using mouse L1210 and the cell nucleus, creating a system for con- P388 leukaemias. Although reproducible, stable and relatively inexpensive, the use of rapidly trolling cell proliferation and cell death. By dividing haematological mouse tumours introduced bias in the screens in favour of agents with the early 1990s, an explosion of drug targets activity against tumours with high growth fractions82.The inadequacy of these screening models transformed cancer drug development from for selecting agents active against solid tumours was implicated, at least in part, for the relatively slow progress in advancing treatments for common tumours throughout the 1960s and 1970s. In a low-budget, government-supported an attempt to find drugs active against solid tumours, the NCI adopted, in 1976, human tumour research effort to a high-stakes, multi-billion xenografts into its in vivo screening programme83. The human tumour xenografts were dollar industry (BOX 2). established either by direct implantation of patient biopsy material or by inoculation of Innovations in technology increased the continuous human tumour cell lines into immunodeficient mice. The first three human tumour success of finding inhibitors of specific targets. xenografts included colon (CX-1), breast (MX-1) and lung (LX-1) tumours, but overall more than Combinatorial chemistry provided thousands 300 xenografts have been established representing most main tumour types. In a parallel effort, of unique structures for in vitro screening for the NCI introduced, in 1989, what was initially called ‘disease-oriented’ screening. This new inhibitors. Molecules identified in high- approach used a rationally designed screening panel containing 60 cell lines derived from seven throughput screens could then be optimized different human cancer types, including colon, brain, lung, melanoma, ovarian, renal and for other properties, including greater speci- leukaemia. Subsequently, breast and prostate cell lines were added84.The use of the human cell ficity or bioavailability. The characteristics of line screening approach was increased by Kenneth Paull and colleagues, who established the promising anticancer drugs became clear — COMPARE algorithm and demonstrated that the growth-inhibitory patterns of anticancer drugs an agent should be metabolically stable, with a against the cell lines correlated well with their mechanism of action57.Unfortunately, none of long half-life in model systems and in 53 these screening systems has successfully predicted outcome of clinical trials .More recent efforts humans, and a slow rate of metabolism by are concentrating on using high-throughput molecular screens followed by the cell-line screens to enzymes such as the cytochrome-P450 family. 82 analyse drug effects . Second, the candidate molecule should be well-absorbed after oral administration, which was not a typical characteristic of the enthusiasm from clinicians, patients and Both drugs continue in clinical trials chemotherapy drugs discovered in the 1970s Congress, who all advocated for new types supported by the NCI; analogues of gel- and 1980s. Finally, it should show a favourable of agent. The screening approaches were danomycin and second-generation cell- toxicity profile at biologically effective doses, not yielding groundbreaking discoveries. cycle inhibitors have also entered clinical with limited effects on bone marrow and There was strong sentiment across the trials and might hold greater promise. intestinal epithelium. board against continuing a screening sys- Perhaps the most important discoveries One of the landmark events in the tar- tem that tested random chemicals and forthcoming from this human cell line geted revolution has been the development natural products against mouse tumours. screen were improvements in screening of imatinib mesylate (Glivec), a relatively In response, the NCI’s Division of methodology, now widely adopted by simple structure that possesses all the Cancer Treatment and its advisors adopted industry, including a rapid colorimetric desired factors of the ‘ideal’ targeted com- a screen based on testing against a panel of assay for cell viability (the MTT assay)56, pound. It was derived from a natural prod- 60 human tumour cell lines, covering a informatic techniques that can identify pat- uct by chemists at Novartis. Imatanib is a broad range of tumour types (BOX 1)53.In terns of cytotoxic response and resistance moderately potent inhibitor of the kinase view of the unique chemistry of compounds among cell lines57,and high-throughput BCR–ABL, the fusion protein product of a found in plants and marine organisms, and automated screening. chromosomal translocation that is involved the important new agents previously dis- in the pathogenesis of chronic myeloid covered from nature, greater emphasis was The targeted-therapy revolution leukaemia (CML). Imatinib also inhibits placed on collecting new species worldwide While the attempts to improve the pace of the KIT tyrosine kinase and platelet derived and on testing their extracts in screens of discovery of cytotoxic agents proceeded in growth factor receptor-β (PDGFRβ) tyro- human tumour cell lines. In hindsight, this the late 1980s, molecular and genetic sine kinase. These latter effects have been screening approach has not been much approaches to understanding cell biology successfully exploited for therapy of gas- more successful in identifying new anti- uncovered entirely new signalling networks trointestinal stromal tumours and the cancer drugs. Although the new cell lines that regulate cellular activities such as prolif- hypereosinophilic syndrome, respectively. used in the screen led to the identification of eration and survival. Many of these net- Brian Druker showed that when imatinib is an increased number of anticancer agents, works were found to be radically altered in used to treat patients with chronic-phase most of these were antimitotic agents cancer cells. An industrial revolution CML, 90% achieve complete haematologi- and topoisomerase I and II inhibitors, much unfolded, based primarily in small biotech- cal remission and many lose cytogenetic like those identified in the original screens. nology firms, as researchers set out to repair evidence of the malignant clone. However, Occasionally, a new class of tumour these molecular defects in cancer cells, BCR–ABL translocation can still be inhibitor developed. For example, gel- beginning the era of ‘targeted therapy’. detected by PCR analysis in cells of most danomycin inhibits heat-shock protein 90 The new targets included growth factors, patients59–61.In the acute leukaemic phase (REF. 54),which is important for regulating signalling molecules, cell-cycle proteins, of CML, imatinib induces brief remissions, protein degradation, and flavopiridol modulators of apoptosis and molecules that and treatment leads to a rapid outgrowth of inhibits a cell-cycle-dependent kinase55. promoted angiogenesis58.Cell biologists drug-resistant cells that display mutations

NATURE REVIEWS | CANCER VOLUME 5 | JANUARY 2005 | 69 PERSPECTIVES in the catalytic kinase domain of ABL. either mutations or in-frame deletions, that development of new drugs71.Otherwise we These mutations hold the enzyme in an are associated with response to therapy in will continue the financially, scientifically and open configuration that binds the drug patients with lung cancer69,70.These receptor ethically unacceptable practice of treating poorly but remains catalytically active62.In mutations and deletions confer a more many patients to benefit a few. some patients in chronic phase CML, drug- robust response to the and a corre- Along with the successes, there have also resistant cells are present before drug expo- sponding increase in sensitivity to receptor been disappointments in the targeted-therapy sure63,a finding usually associated with inhibition by gefitinib. Alterations in EGFR revolution. A series of farnesyl transferase rapid emergence of resistance64.This find- are found in 10% of patients with NSCLC in inhibitors (FTIs), developed and tested ing proves that cancers, through their the United States, but, interestingly, in 25% against HRAS,entered the clinic in the late intrinsic mutability, contain a range of or more of patients with the same disease in 1990s but failed to cause regression of human drug-resistant mutant subclones, even Japan, where the response rate to gefitinib is solid tumours72,73.Some FTIs, however, before treatment. correspondingly higher. Most responders to remain in clinical trails against acute myeloid A second class of compounds, those that therapy are women and non-smokers, and leukaemia and myelodysplasia. The reason for inhibit the epidermal growth factor receptor their tumours often contain elements of their lack of efficacy against solid tumours is (EGFR), has won FDA approval, but with bronchoalveolar carcinoma, indicating that not obvious. The FTIs clearly affect proteins less marked evidence of antitumour activity. these patients constitute a distinct subset that are subject to farnesylation, but they are Gefitinib (Iressa), a competitive inhibitor of within the larger histological classification not consistently effective against KRAS-trans- the ATP-binding function of the tyrosine- of NSCLC. It is possible that the mutation is formed cells in culture. Alternative pathways kinase active site of EGFR, causes partial a transforming event in these patients, and for KRAS lipid modulation might circumvent remissions in 10–15% of patients with non- that the tumour is in essence ‘addicted’ to the block. small-cell lung cancer (NSCLC)65,but failed the receptor’s activation. A second high-profile failure added to the to increase activity of chemotherapy in large The discovery of the mutations that initial gloom of the new age of biotechnology, randomized trials66,67.A monoclonal anti- confer sensitivity to EGFR inhibitors repre- but has more recently been vindicated by suc- body against the extracellular domain of sents an important advance in the strategy cessful clinical trials. In the early 1970s, Judah EGFR, cetuximab (Erbitux), also won FDA of cancer drug development. Most agents in Folkman demonstrated for the first time the approval in 2003, in this case in combina- current use, including cytotoxics and the central role of angiogenesis in allowing tumour tion with chemotherapy for colon cancer68. newer targeted drugs, have relatively low proliferation and metastasis74.Folkman, It is not clear that the two agents, which tar- response rates in patients with any given Harold Dvorak and others found that tumour get the same receptor in quite different histological class of solid tumours. The use cells secreted angiogenic molecules such as ways, produce responses in the same sub- of molecular or genomic tests, or pro- PDGF and VEGF,which stimulated new vessel set of patients or the same diseases. teomics, to identify reliable markers for formation in their environs and made available Subsequently, researchers at Harvard have drug sensitivity will undoubtedly become a supply of nutrients that allowed further identified molecular changes in EGFR, one of the key objectives early in the clinical expansion of the tumour. Despite the clinical failure of the anti-angiogenic peptide endo- statin, several small molecules (SU-11248, Box 2 | The growth of chemotherapy as an industry Bayer 43-9006) that inhibit the VEGF receptor 2, and an anti-VEGF antibody, bevacizumab The drugs and biological agents that oncologists use to treat cancer now amount to a multi- (Avastin), have clear antitumour activity in billion dollar industry. In 2003, the total worldwide market for oncology-related products was approximately US $36.8 billion. Today, expenditures on oncology-related products represent patients with renal-cell carcinoma, a tumour almost 10% of the total $430 billion worldwide market for pharmaceuticals, and the oncology that arises through mutation or loss of a 75 market is projected to exceed $60 billion by 2008 (REF. 85).Cancer-related chemotherapy drugs functional VHL gene . VHL suppresses the and biological treatments comprise approximately 48% and 15% of the total market size, function of hypoxia-inducible factor-1α,the respectively, whereas haematological growth factors, used mainly to support the use of transcription factor that responds to hypoxia chemotherapy, account for the remaining 37% of the market. A combination of factors has and stimulates angiogenesis. Bevacizumab also driven the growth in the oncology market, including expanded therapeutic options, greater increases the activity of standard cytotoxic willingness of oncologists to treat older patients, and relatively high prices of many of the drugs against colon cancer and achieved FDA newer anticancer agents. The cost of drugs to treat colon cancer is illustrative. Before 1996, approval in 2004 for treating patients with this 5-fluorouracil (5-FU) was the principle agent used to treat advanced colon cancer. The drug cost cancer76. SU-11248, which also inhibits KIT,is for 8 weeks of treatment with a regimen based on 5-FU was under $100. By 2004, the Food and also active in patients with gastrointestinal Drug Administration had approved 5 additional agents for the treatment of colon cancer, stromal tumours, including those refractory including irinotecan, , , cetuximab and bevacizumab. The drug cost for to imatinib. the same 8 weeks of treatment can now exceed $30,000, as the newer drugs have been added 86 onto, rather than replaced, existing agents . The size and scope of the enterprise focused on The past as prologue cancer drug development has also grown. Cancer drug development has transformed from a From these experiments it became clear that small, mostly public effort focused in the United States to become a major international the path to success for targeted molecules industrial effort. At present, more than 1,300 small biotech companies have formed in the United would be no less arduous than the trials for States alone to develop products based on molecular targets87,and more than half are focusing cytotoxic drugs. Certain principles emerged on treating cancer. According to a recent industry-wide survey, there are now at least 395 agents from the cytotoxic era, and are likely to be for the treatment of cancer in clinical trials88,more than in any other therapeutic class of applied to the new chemotherapies. First, ani- medicine89.The pharmacoeconomic challenges of paying for all of these new agents will almost certainly become an area of increased attention as the oncology field moves forward. mal models, while instructive, are unreliable predictors for success against human disease.

70 | JANUARY 2005 | VOLUME 5 www.nature.com/reviews/cancer PERSPECTIVES

30 earlier. Ironically, the scenario of evolution treatment. Additional major scientific chal- of drug resistance has been most elegantly lenges for the next generation will involve 25 studied for two successful agents, methotrex- exploiting the mutability of cancer cells and ate (the cytotoxic folate analogue) and ima- reversing their resistance to apoptosis80.Will 20 tinib (the highly effective inhibitor of the medical treatments provide cures for most BCR–ABL tyrosine kinase). In some cases, common cancers over the next 60 years? 15 targeted drugs and conventional antitumour They will probably do so for subsets of the agents can both be affected by a common main tumour types, and will undoubtedly 10

Number of approvals resistance mechanism involving drug efflux extend survival and improve quality of life (the MDR transporter)77,or mutations in for many others. 5 cell-death pathways. Because of resistance to Bruce A. Chabner and Thomas G. Roberts Jr are at single agents, combination therapy is essen- the Division of Hematology/Oncology, 0 tial for tumour eradication and cure. As we Massachusetts General Hospital, Harvard Medical have reviewed, even the most successful of School, Boston, Massachusetts, 02114 USA. the targeted molecules, imatinib, does not fully Correspondence to B.A.C. 1971–19751976–19801981–19851986–19901991–19951996–2000 e-mail: [email protected] 2001–present eradicate the malignant clone. Fortunately, doi:10.1038/nrc1529 Figure 5 | Number of approved new molecules clinical trials have demonstrated potent syn- for the treatment of cancer by the Food and ergy between targeted molecules, particularly 1. Papac, R. J. Origins of cancer therapy. Yale J. Biol. Med. Drug Administration. The number drugs 74, 391–398 (2001). approved for the treatment of cancer by the United monoclonal antibodies such as rituximab 2. Gilman, A. The initial clinical trial of nitrogen mustard. Am. States Food and Drug Administration (FDA) per 5- (Rituxan)78,bevacizumab76 and trastuzumab J. Surg. 105, 574–578 (1963). year period was relatively constant through 1990, 79 3. Gilman, A. The biological actions and therapeutic (Herceptin) ,and traditional chemotherapy. applications of the B-chloroethyl amines and sulfides. and has increased substantially since then. Data It has become increasingly clear that Science 103, 409–436 (1946). are updated through August 2004 and are available understanding the molecular profile of 4. Wills, L., Clutterbuch, P. & Evans, B. D. F. A new factor in from the FDA web site (see online links box). the production and cure of macrocytic anaemias and its human tumours is essential to the effective relation to other haemopoietic principles curative in pernicious anaemia. Biochem. J. 31, 2136–2147 (1937). use of cancer drugs. For 60 years, clinicians 5. Farber, S., Diamond, L. K., Mercer, R. D., Sylvester, R. F. have depended on histological classification & Wolff, J. A. Temporary remissions in acute in children produced by folic antagonist, 4- Human tumours of any given histological of tumours to dictate therapeutic choices, aminopteroylglutamic acid (aminopterin). N. Engl. J. Med. type have great genetic diversity, as revealed despite the fact that most patients with a solid 238, 787–793 (1948). 6. Bonadonna, G. et al. Combination chemotherapy as an by gene-expression profiling, and in most tumour do not respond to any given drug. adjuvant treatment in operable breast cancer. N. Engl. J. types of cancer only a subset of patients will Patient selection, based on molecular features Med. 294, 405–410 (1976). 7. Li, M. C., Hertz, R. & Bergenstal, D. M. Therapy of prove responsive to any given new agent. of the tumour, confers enormous advantage choriocarcinoma and related trophoblastic tumors with Animals imperfectly reflect the pharmaco- in trial design, allows for more efficient and folic acid and purine antagonists. N. Engl. J. Med. 259, kinetics of drugs in humans, because of cost-effective drug development, and ulti- 66–74 (1958). 8. Jaffe, N., Frei, E. 3rd, Traggis, D. & Bishop, Y. Adjuvant more rapid metabolism, greater tolerance mately will reduce the burgeoning cost of new methotrexate and citrovorum-factor treatment of for side effects and differences in protein technology. Molecular profiling and the study osteogenic sarcoma. N. Engl. J. Med. 291, 994–997 (1974). binding. Even specific organ toxicities can- of patient selection must become a central 9. Jaffe, N. et al. High-dose methotrexate in osteogenic not be extrapolated from mice to humans. aim of cancer drug development71. sarcoma. Natl Cancer Inst. Monogr. 56, 201–206 (1981). 10. Osborn, M. J., Freeman, M. & Huennekens, F. M. Inhibition As pointed out recently by Douglas The discovery and development of cancer of dihydrofolic reductase by aminopterin and amethopterin. Hanahan and Robert Weinberg58, the basic chemotherapy has therefore come full circle. Proc. Soc. Exp. Biol. Med. 97, 429–431 (1958). 11. Osborn, M. J. & Huennekens, F. M. Enzymatic reduction biology of murine cells, and the process of We have more and better drugs (FIG. 5) and of dihydrofolic acid. J. Biol. Chem. 233, 969–974 (1958). transformation, can differ significantly improved knowledge of cancer as a disease, 12. Jolivet, J., Cowan, K. H., Curt, G. A., Clendeninn, N. J. & Chabner, B. A. The pharmacology and clinical use of from that of human cells. Attempts to pro- but the early observations of Goodman, methotrexate. N. Engl. J. Med. 309, 1094–104 (1983). duce genetically engineered mouse models Gilman and Farber remain accurate and, 13. Messmann, R. A. & Allegra, C. J. in Cancer Chemotherapy and Biotherapy (eds Chabner, B. A. & of human cancer in fact lead to models of indeed, immutable. The transition from Longo, D.) 139–184 (Lippincott Williams & Wilkins, the specific molecular changes in a mouse cytotoxic drugs to targeted therapies repre- Philadelphia, 2001). 14. Curt, G. A., Clendeninn, N. J. & Chabner, B. A. Drug cell, and have uncertain relevance to a sents an important advance, but the basic resistance in cancer. Cancer Treat. Rep. 68, 87–99 (1984). human counterpart. principles of cancer treatment and drug 15. Stoller, R. G., Hande, K. R., Jacobs, S. A., Rosenberg, S. A. & Chabner, B. A. Use of plasma pharmacokinetics to A second enduring principle has been resistance, as developed in the period from predict and prevent methotrexate toxicity. N. Engl. J. that human tumours, when they do 1950 to 1980, remain the same. Human Med. 297, 630–634 (1977). 16. Skipper, H. E., Thomson, J. R., Elion, G. B. & Hitchings, G. H. respond, contain subclones that become malignancies are a very diverse group of dis- Observations on the anticancer activity of drug resistant. Under the selective pressure eases, even within histological classifications, 6-mercaptopurine. Cancer Res. 14, 294–298 (1954). 17. Hitchings, G. H. & Elion, G. B. The chemistry and of a toxic therapy, the genetic diversity and quickly display their diversity when biochemistry of purine analogs. Ann. NY Acad. Sci. 60, within most human tumours leads to rapid exposed to all forms of chemotherapy. The 195–199 (1954). 18. Johnson, I. S., Armstrong, J. G., Gorman, M. & Burnett, J. P. Jr. outgrowth of drug-resistant cells. A vast next decade will present the challenge of The Vinca alkaloids: a new class of oncolytic agents. array of resistance mechanisms, involving designing trials to combine targeted drugs Cancer Res. 23, 1390–1427 (1963). 19. Bensch, K. G. & Malawista, S. E. Microtubule crystals: a mutations or amplification of the target and cytotoxics in a more effective manner. new biophysical phenomenon induced by Vinca enzyme, overexpression of drug trans- These trials will be aided by the use of alkaloids. Nature 218, 1176–1177 (1968). 20. Frei, E. 3rd et al. The effectiveness of combinations of porters, or mutations in cell-death pathways, genomic and molecular assays to identify antileukemic agents in inducing and maintaining can defeat single agents, no matter how well subsets of patients that are most likely to remission in children with acute leukemia. Blood 26, 642–656 (1965). designed and targeted — this was also respond to certain drugs, thereby avoiding 21. Frei, E. 3rd. The National Cancer Chemotherapy observed by Gilman and colleagues 60 years the needless cost and toxicity of ineffective Program. Science 217, 600–606 (1982).

NATURE REVIEWS | CANCER VOLUME 5 | JANUARY 2005 | 71 PERSPECTIVES

22. Driscoll, J. S. The preclinical new drug research program 48. Pizzo, P. A. Granulocytopenia and cancer therapy. Past 73. Sharma, S. et al. A phase II trial of farnesyl protein of the National Cancer Institute. Cancer Treat. Rep. 68, problems, current solutions, future challenges. Cancer transferase inhibitor SCH 66336, given by twice-daily oral 63–76 (1984). 54, 2649–2661 (1984). administration, in patients with metastatic colorectal 23. Grever, M. R., Schepartz, S. A. & Chabner, B. A. The 49. Lieschke, G. J. & Burgess, A. W. Granulocyte colony- cancer refractory to 5-fluorouracil and irinotecan. Ann. National Cancer Institute: cancer drug discovery and stimulating factor and granulocyte-macrophage colony- Oncol. 13, 1067–1071 (2002). development program. Semin. Oncol. 19, 622–638 stimulating factor (1). N. Engl. J. Med. 327, 28–35 (1992). 74. Folkman, J. Tumor angiogenesis: therapeutic (1992). 50. Lieschke, G. J. & Burgess, A. W. Granulocyte colony- implications. N. Engl. J. Med. 285, 1182–1186 (1971). 24. Skipper, H. E., Schabel, F. M. Jr. & Wilcox, W. S. stimulating factor and granulocyte-macrophage colony- 75. Gnarra, J. R. et al. Mutations of the VHL tumour Experimental evaluation of potential anticancer agents. stimulating factor (2). N. Engl. J. Med. 327, 99–106 (1992). suppressor gene in renal carcinoma. Nature Genet. 7, Xiii. On the criteria and kinetics associated with ‘curability’ 51. Curtis, R. E. et al. Risk of leukemia after chemotherapy 85–90 (1994). of experimental leukemia. Cancer Chemother. Rep. 35, and radiation treatment for breast cancer. N. Engl. J. 76. Hurwitz, H. et al. Bevacizumab plus irinotecan, 1–111 (1964). Med. 326, 1745–1751 (1992). fluorouracil, and leucovorin for metastatic colorectal 25. Skipper, H. E. & Griswold, D. P. Frank M. Schabel 52. Burstein, H. J. & Winer, E. P. Primary care for survivors of cancer. N. Engl. J. Med. 350, 2335–2342 (2004). 1918–1983. Cancer Res. 44, 871–872 (1984). breast cancer. N. Engl. J. Med. 343, 1086–1094 (2000). 77. Roninson, I. B. et al. Isolation of human mdr DNA 26. Moxley, J. H. 3rd, De Vita, V. T., Brace, K. & Frei, E. 3rd. 53. Johnson, J., Monks, A., Hollingshead, M. & Sausville, E. sequences amplified in multidrug-resistant KB carcinoma Intensive combination chemotherapy and X-irradiation in in Cancer Chemotherapy and Biotherapy (eds Chabner, cells. Proc. Natl Acad. Sci. USA 83, 4538–4542 (1986). Hodgkin’s disease. Cancer Res. 27, 1258–1263 (1967). B. A. & Longo, D.) 17–36 (Lippincott Williams & Wilkins, 78. Coiffier, B. et al. CHOP chemotherapy plus rituximab 27. Devita, V. T. Jr., Serpick, A. A. & Carbone, P. P. Philadelphia, 2001). compared with CHOP alone in elderly patients with Combination chemotherapy in the treatment of advanced 54. Whitesell, L., Mimnaugh, E. G., De Costa, B., Myers, C. E. diffuse large-B-cell lymphoma. N. Engl. J. Med. 346, Hodgkin’s disease. Ann. Intern. Med. 73, 881–895 & Neckers, L. M. Inhibition of heat shock protein HSP90- 235–242 (2002). (1970). pp60v-src heteroprotein complex formation by 79. Slamon, D. J. et al. Use of chemotherapy plus a 28. Levitt, M. et al. Combination sequential chemotherapy in benzoquinone ansamycins: essential role for stress monoclonal antibody against HER2 for metastatic breast advanced reticulum cell sarcoma. Cancer 29, 630–636 proteins in oncogenic transformation. Proc. Natl Acad. cancer that overexpresses HER2. N. Engl. J. Med. 344, (1972). Sci. USA 91, 8324–8328 (1994). 783–792 (2001). 29. Berd, D., Cornog, J., DeConti, R. C., Levitt, M. & Bertino, J. R. 55. Worland, P. J. et al. Alteration of the phosphorylation 80. Reed, J. C. Dysregulation of apoptosis in cancer. J. Clin. Long-term remission in diffuse histiocytic lymphoma state of p34cdc2 kinase by the flavone L86-8275 in breast Oncol. 17, 2941–2953 (1999). treated with combination sequential chemotherapy. carcinoma cells. Correlation with decreased H1 kinase 81. Zubrod, C. G. The national program for cancer Cancer 35, 1050–1054 (1975). activity. Biochem. Pharmacol. 46, 1831–1840 (1993). chemotherapy. JAMA 222, 1161–1162 (1972). 30. Moertel, C. G. et al. Levamisole and fluorouracil for 56. Mosmann, T. Rapid colorimetric assay for cellular growth 82. Takimoto, C. H. Anticancer drug development at the US adjuvant therapy of resected colon carcinoma. N. Engl. J. and survival: application to proliferation and cytotoxicity National Cancer Institute. Cancer Chemother. Med. 322, 352–358 (1990). assays. J. Immunol. Methods 65, 55–63 (1983). Pharmacol. 52 (Suppl. 1), 29–33 (2003). 31. McGuire, W. P. et al. Taxol: a unique antineoplastic agent 57. Paull, K. D. et al. Display and analysis of patterns of 83. Kelland, L. R. Of mice and men: values and liabilities of with significant activity in advanced ovarian epithelial differential activity of drugs against human tumor cell the athymic nude mouse model in anticancer drug neoplasms. Ann. Intern. Med. 111, 273–279 (1989). lines: development of mean graph and COMPARE development. Eur. J. Cancer 40, 827–836 (2004). 32. Goodman, J. & Walsh, V. The Story of Taxol: Nature and algorithm. J. Natl Cancer Inst. 81, 1088–1092 (1989). 84. Boyd, M. R. in Anticancer Drug Development Guide: Politics in the Pursuit of an Anti-cancer Drug (Cambridge 58. Hanahan, D. & Weinberg, R. A. The hallmarks of cancer. Preclinical Screening, Clinical Trials, and Approval (ed. Univ., Cambridge, 2001). Cell 100, 57–70 (2000). Teicher, B.) 23 (Humana, Totowa, 1997). 33. Saltz, L. B. et al. Irinotecan plus fluorouracil and 59. Druker, B. J. et al. Efficacy and safety of a specific inhibitor 85. S. G. Cowen & Company. Pharmaceutical Therapeutic leucovorin for metastatic colorectal cancer. Irinotecan of the BCR–ABL tyrosine kinase in chronic myeloid Categories Outlook: Comprehensive Study 375 (S.G. Study Group. N. Engl. J. Med. 343, 905–914 (2000). leukemia. N. Engl. J. Med. 344, 1031–1037 (2001). Cowen & Company, New York, 2004). 34. Noda, K. et al. Irinotecan plus cisplatin compared with 60. Kantarjian, H. et al. Hematologic and cytogenetic 86. Schrag, D. The price tag on progress — chemotherapy plus cisplatin for extensive small-cell lung responses to imatinib mesylate in chronic myelogenous for colorectal cancer. N. Engl. J. Med. 351, 317–319 cancer. N. Engl. J. Med. 346, 85–91 (2002). leukemia. N. Engl. J. Med. 346, 645–652 (2002). (2004). 35. Bodurka, D. C. et al. Phase II trial of irinotecan in patients 61. Hughes, T. P. et al. Frequency of major molecular 87. Ernst & Young LLP, Annual Biotechnology Industry with metastatic epithelial or peritoneal responses to imatinib or interferon alfa plus in Report, 2001. In Parexel’s Pharmaceutical R&D Statistical cancer. J. Clin. Oncol. 21, 291–297 (2003). newly diagnosed chronic myeloid leukemia. N. Engl. J. Sourcebook (Parexel, Walthan, 2002/2003). 36. Von Hoff, D. D. There are no bad anticancer agents, only Med. 349, 1423–1432 (2003). 88. Pharmaceuticals Research and Manufacturers of bad clinical trial designs — twenty-first Richard and 62. Shah, N. P. et al. Multiple BCR–ABL kinase domain America. New Medicines in Development for Cancer: 395 Hinda Rosenthal Foundation Award Lecture. Clin. Cancer mutations confer polyclonal resistance to the tyrosine New Medicines in Development Offer Hope in the War on Res. 4, 1079–1086 (1998). kinase inhibitor imatinib (STI571) in chronic phase and Cancer 1–56 (PhRMA, Washington, 2003). 37. Johnson, J. I. et al. Relationships between drug activity in blast crisis chronic myeloid leukemia. Cancer Cell 2, 89. IMS LifeCycle, R&D focus. In Parexel’s Pharmaceutical R&D NCI preclinical in vitro and in vivo models and early 117–125 (2002). Statisitcal Sourcebook (Parexel, Walthan, 2003/2004). clinical trials. Br. J. Cancer 84, 1424–1431 (2001). 63. Roche-Lestienne, C. et al. Several types of mutations of 90. Chabner, B. The traveling oncologist and the wages of 38. Rosenberg, B., Vancamp, L. & Krigas, T. Inhibition of cell the Abl gene can be found in chronic myeloid leukemia sin. Oncologist 6, 1–2 (2001). division in Escherichia coli by electrolysis products from a patients resistant to STI571, and they can pre-exist to the 91. Gorlick, R. et al. Intrinsic and acquired resistance to platinum electrode. Nature 205, 698–699 (1965). onset of treatment. Blood 100, 1014–1018 (2002). methotrexate in acute leukemia. New Engl. J. Med. 335, 39. Bosl, G. J. et al. VAB-6: an effective chemotherapy 64. Branford, S. et al. Detection of BCR–ABL mutations in 1041–1048 (1996). regimen for patients with germ-cell tumors. J. Clin. patients with CML treated with imatinib is virtually Oncol. 4, 1493–1499 (1986). always accompanied by clinical resistance, and Competing interests statement 40. Evans, B. D., Raju, K. S., Calvert, A. H., Harland, S. J. & mutations in the ATP phosphate-binding loop (P-loop) The authors declare competing financial interests: see Web Wiltshaw, E. Phase II study of JM8, a new platinum are associated with a poor prognosis. Blood 102, version for details. analog, in advanced ovarian carcinoma. Cancer Treat. 276–283 (2003). Rep. 67, 997–1000 (1983). 65. Kris, M. G. et al. Efficacy of gefitinib, an inhibitor of the 41. Montgomery, J. A. Chemistry and structure-activity epidermal growth factor receptor tyrosine kinase, in Online links studies of the . Cancer Treat. Rep. 60, symptomatic patients with non-small cell lung cancer: a 651–664 (1976). randomized trial. JAMA 290, 2149–2158 (2003). DATABASES 42. Walker, M. D. & Hurwitz, B. S. BCNU (1,3-bis(2- 66. Giaccone, G. et al. Gefitinib in combination with The following terms in this article are linked online to: chloroethyl)-1-; NSC-409962) in the treatment and cisplatin in advanced non-small-cell Entrez Gene: of malignant brain tumor — a preliminary report. Cancer lung cancer: a phase III trial — INTACT 1. J. Clin. Oncol. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene Chemother. Rep. 54, 263–271 (1970). 22, 777–784 (2004). ABL | BCR | EGFR | HRAS | KIT | KRAS | PDGFRβ | VHL 43. Wilson, C. B., Boldrey, E. B. & Enot, K. J. 1,3-bis (2- 67. Herbst, R. S. et al. Gefitinib in combination with paclitaxel National Cancer Institute: http://cancer.gov/ chloroethyl)-1-nitrosourea (NSC-409962) in the treatment and in advanced non-small-cell lung cancer: acute lymphoblastic leukaemia | acute myeloid leukaemia | of brain tumors. Cancer Chemother. Rep. 54, 273–281 a phase III trial — INTACT 2. J. Clin. Oncol. 22, 785–794 bladder cancer | breast cancer | chronic myeloid leukaemia | (1970). (2004). colon cancer | head and neck cancer | Hodkin’s lymphoma | 44. Rai, K. R. et al. compared with chlorambucil 68. Cunningham, D. et al. Cetuximab monotherapy and non-Hodgkin’s lymphoma | non-small-cell lung cancer | ovarian as primary therapy for chronic lymphocytic leukemia. cetuximab plus irinotecan in irinotecan-refractory cancer | testicular cancer N. Engl. J. Med. 343, 1750–1757 (2000). metastatic colorectal cancer. N. Engl. J. Med. 351, 45. Minocha, A. & Long, B. H. Inhibition of the DNA 337–345 (2004). FURTHER INFORMATION catenation activity of type II topoisomerase by VP16-213 69. Lynch, T. J. et al. Activating mutations in the epidermal Developmental Therapeutics Program NCI/NIH: and VM26. Biochem. Biophys. Res. Commun. 122, growth factor receptor underlying responsiveness of non- http://dtp.nci.nih.gov 165–170 (1984). small-cell lung cancer to gefitinib. N. Engl. J. Med. 350, FDA’s Oncology Tools web site (including approval 46. Freireich, E. J., Schmidt, P. J., Schneiderman, M. A. & 2129–2139 (2004). statistics): http://www.fda.gov/cder/cancer/approved.htm Frei, E. 3rd. A comparative study of the effect of 70. Paez, J. G. et al. EGFR mutations in lung cancer: NCI’s Cancer Therapy Evaluate Program: transfusion of fresh and preserved whole blood on correlation with clinical response to gefitinib therapy. http://ctep.cancer.gov/ bleeding in patients with acute leukemia. N. Engl. J. Med. Science 304, 1497–1500 (2004). NCI’s Closing in on Cancer web site: 260, 6–11 (1959). 71. Roberts, T. G. Jr. & Chabner, B. A. Beyond fast track for http://press2.nci.nih.gov/sciencebehind/cioc/ciocframe.htm 47. Gaydos, L. A., Freireich, E. J. & Mantel, N. The drug approvals. N. Engl. J. Med. 351, 501–505 (2004). The American Cancer Society History of Cancer web site: quantitative relation between platelet count and 72. Adjei, A. A. et al. Phase II study of the farnesyl transferase http://www.cancer.org/docroot/CRI/content/CRI_2_6x_the_hi hemorrhage in patients with acute leukemia. N. Engl. J. inhibitor R115777 in patients with advanced non-small- story_of_cancer_72.asp?sitearea= Med. 266, 905–909 (1962). cell lung cancer. J. Clin. Oncol. 21, 1760–1766 (2003). Access to this interactive links box is free online.

72 | JANUARY 2005 | VOLUME 5 www.nature.com/reviews/cancer